Every now and then I get a question I just can’t answer, which is great because it gives us both an opportunity to learn. If anyone can help with this, I’d love to hear your thoughts. Please email me.
Here’s the email Q&A…
Hi again Lance,
I have a real question to ask as I really don’t know the answer and have been pondering it as one of my obsessions with trying to understand the market/s.
The question surrounds the basis for support and resistance prices but as I think about it, it can also apply to trend lines as well. And specifically, I am thinking about indices. I understand the notion of support and resistance with individual stock, futures and currency trading. They are unitary in that they are not comprised of any other combination of securities. However, thinking about an index like the S&P 500 or the Dow is a bit of a conundrum.
Consider: An index like the Dow or the S&P are merely a number of individual components mathematically weighted and summed to create a resulting value. In the case of the S&P there are ~ 2700 issues in this mathematical total called the S&P 500. Yet, one looks at these indexes as if they were a singular entity when in fact they are only a summation of the many individual elements. How is it possible and what is the mathematical basis for 2700 issues converging exactly in such a manner as to create a specific resistance or support level in the index given that each is only a component of the index and not all traders/investors, trade all of the issues in exactly the same proportion at the same time? Support and resistance levels may well apply to an individual issue but somehow, it doesn’t make sense that they all conspire to create specific and repeating support and resistance levels in the index to which they are only a part.
I hope what I expressed makes sense. It just is one of those fundamental understandings that have been bothering me and that I believe can apply to a better understanding as to why the support and resistance levels seem to appear consistently on the indexes. Of course, you can trade the DIA which is just a 100th of the DOW but it’s only a divided value of the DOW and thus inherits the same index properties. The same question applies here.
I have never heard this issue discussed or illuminated and I can find no discussion on the internet. Any thoughts?
How do support and resistance form in indices? Great question.
Unfortunately I’m not sure I can provide an equally great answer. I’m happy to share my take on the issue & maybe it’ll give you some insight which can then help to further my understanding.
S/R exists on tradable instruments such as stocks, future, currencies, due to the nature of human decision making. Primarily a result of anchoring bias although there are other factors involved as well . Anchoring creates S/R due to our need to reference current price against some other price in order to assess current value as good value or poor value. The anchors usually chosen are those key points that stand out on charts – previous swing highs/lows, gaps, areas of congestion (and in particular the point of breakout). There’s a whole lot more to it, but that’s the summary version (see behavioural finance studies, in particular heuristics & biases, for more).
So how can this concept apply to a composite index, as these charts also clearly show S/R levels.
I don’t believe it does apply. I suspect there’s different factors involved although I haven’t really had enough thought on this topic.
Here’s a couple of ideas though.
Firstly, save and then open the excel spreadsheet available from this link. This is a random number generator which displays its output as a series of 200 price bars, and their corresponding RSI. You’ll see the rules it uses for construction of the price bars above the chart display. Press F9 as many times as you wish to see further charts, once again based on a new sequence of random numbers. (I’m not sure of the influence the programming of the excel random number function would have on validity of data – I’m operating on the assumption that it’s close enough to random).
The interesting observation from this is that random data when plotted in price chart format appear quite clearly to show characteristics of financial charts. Pressing F9 multiple times and you’ll see all kind of chart formations – trends, support & resistance, head and shoulder patterns, triangles, double tops / bottoms etc.
This is not evidence of the random walk concept. The fact that random data appears to share the same characteristics as financial charts, does not mean that financial charts are random. In fact, I’m told that when comparing plots of the standard deviation of price closes vs the same for random data, the result is quite different. Random data forms a nice normalized bell curve. The financial chart data apparently is more narrow in the body with a very wide base, indicating greater periods of consolidation which are separated by larger than random impulse move. So I’m told anyway – I don’t have any proof of this and am in fact not a statistician (apologies as well for any incorrect terminology). The point is though, financial markets are not random.
What this might mean though, is simply that our mind is very good at seeing patterns in data.
There are no S/R levels on non-tradable indices if we define S/R levels as those created by human decision making. Rather we perceive S/R levels on these charts simply because any data plotted in this format (even random data) exhibits similar characteristics which look like S/R.
That may be a partial answer.
Another option, which possibly plays some part in the presence of S/R on indices, is the degree to which analysis of the index then influences our analysis and trading of the component instruments. Given the prevalence of top-down analysis, where traders start with the wider market analysis, then sector analysis, then individual stocks, is it not possible that this plays a part. For example, if the index is approaching what we perceive as strong resistance and the marketplace is expecting the level to hold, won’t we then be looking for sectors that are highly correlated, and then stocks within that sector exhibiting similar characteristics. Shorting those stocks then creates resistance, not only on that individual stock, but also on both the sector and market index?
If you have any additional ideas, or come across anything on the net, I’d love to hear them.
PS. I’m not sure where the excel random market program came from. I’ve had it for years & don’t recall its source. If anyone knows where, please let me know so I can credit the author & link to their site.